Fintech / Applied AI / Engineering
Dinuwan Fernando
Co-Founder & CTO at Avagance
Building AI-native wealth management infrastructure for the UK financial advisory industry
- 600+
- Service modules
- 200+
- API route registries
- 2024
- Building Avagance
01About
Building the technical core of an AI-native wealth platform
I'm Dinuwan Fernando, Co-Founder and Chief Technology Officer at Avagance. Together with my co-founder Lithika Ranepura, we're building an AI-native wealth management platform designed for UK Independent Financial Advisors and wealth management firms.
I lead the technical architecture and engineering at Avagance, where we've built a TypeScript modular monolith with 600+ service modules, a Python ML microservice, real-time streaming infrastructure, and a multi-tenant data model designed for FCA-regulated financial services.
Before Avagance, I built machine learning models for credit risk prediction, full-stack business applications, and desktop CRM tools. I hold a BSc (Hons) in Software Engineering from the University of Plymouth.
Focus
- Agentic AI systems
- Full-stack TypeScript
- Fintech integrations and compliance
- Applied machine learning
The stack I build with
- TypeScript
- Next.js
- React
- Node.js
- Express
- PostgreSQL
- Redis
- Python
- FastAPI
- PyTorch
- Docker
- Google Cloud Platform
- Drizzle ORM
- Tailwind CSS
02Experience
Leading engineering at Avagance
Co-Founder & Chief Technology Officer
2024 - PresentLeading the technical architecture and engineering of an AI-native wealth management platform for UK IFAs. Built the full-stack infrastructure from the ground up - TypeScript monorepo, Express API with 200+ route registries, Next.js advisor portal, PostgreSQL with Drizzle ORM, Python ML microservice, and GCP deployment pipeline.
Education
BSc (Hons) Software Engineering
University of Plymouth
Delivered through NSBM Green University
03Projects
Selected work
A full-scale wealth management platform for UK financial advisors featuring compliance automation, AI-assisted report generation, real-time portfolio analytics, and multi-tenant architecture.
- TypeScript
- Next.js
- Express
- PostgreSQL
- Redis
- Python
- FastAPI
- GCP
Built a machine learning pipeline comparing Random Forest and XGBoost for mortgage loan approval prediction. Includes feature engineering, model evaluation, and a Tkinter GUI for real-time predictions.
- Python
- XGBoost
- scikit-learn
- pandas
- Jupyter
A modern, responsive business website built with Next.js and TypeScript, integrated with a desktop CRM application for direct customer appointment bookings.
- Next.js
- TypeScript
- Supabase
- Tailwind CSS
Dimuthu Electronics Manager
PrivateAn installable desktop application for managing business operations - appointment scheduling, inventory tracking, receipt generation, and employee management. Integrated with the Dimuthu Electronics website.
- Electron
- TypeScript
04Writing
Notes on building fintech and applied ML
05Contact
Get in touch
Open to conversations about fintech, applied AI, and engineering leadership. The fastest way to reach me is by email.
dinuwan@broinfinance.com